Written evidence for the European Scrutiny Committee

David Collins

Professor of International Economic Law

City, University of London

This submission will address the following question only:

What are the key features of the dispute resolution procedures provided for in the TCA and what are the likely legal and policy implications of these for the UK? How closely do they follow precedent in other trade agreements and do they raise any concerns with respect to the UK’s regulatory autonomy?

This submission will consider the extent to which the TCA’s Level Playing Field (LPF) obligations and associated dispute settlement rules constrain the UK’s capacity to pursue its own regulatory agenda. It suggests that the LPF obligations are narrow and should not pose a significant obstacle to the UK’s plans for regulatory reform.

During the negotiations of their future relationship following the United Kingdom’s departure from the European Union in 2020, the EU insisted that any Free Trade Agreement (FTA) between the parties would need to have “robust guarantees for a level playing field … to ensure open and fair competition among our businesses.”[1] The assurance against unfair competition, achieved via a feared ‘race to the bottom’ in standards was as much of a vital ingredient in the conclusion of successful negotiations between the parties as the UK’s aim to have the regulatory autonomy to unleash marked-based innovations and reforms, one of the primary motivations for the Brexit vote.[2]

Concluded finally in December of 2020, the LPF obligations of the UK-EU Trade and Cooperation Agreement (TCA) require that the parties remain on an equal regulatory footing with a view to preventing unfair competition for foreign investment and trade while acknowledging that the purpose of the agreement is not to ‘harmonise the standards of the parties.’[3] The LPF system is innovative because, as it governs continued relations between the EU and a former member of its Single Market, ultimately the TCA is an FTA uniquely designed to enable divergence rather than convergence. At the same time it aims to facilitate trade and investment between the parties in a manner which ‘stands the test of time.’[4] Thus the LPF commitments in the agreement attempt to straddle a narrow line along which parties are able to veer away from existing rules provided that this does lead to significant, quantifiable harm to trade or investment flows between the parties. It seeks to achieve this while not deterring parties from pursuing their own regulatory reforms since any resulting retaliation by the other party is tightly circumscribed as well as subject to the TCA’s dispute settlement provisions.

              The essence of the TCA’s LFP requirement is contained in the statement found in Article 355.4:

The Parties affirm their common understanding that their economic relationship can only deliver benefits in a mutually satisfactory way if the commitments relating to a level playing field for open and fair competition stand the test of time, by preventing distortions of trade or investment, and by contributing to sustainable development. However, the Parties recognise that the purpose of this Title is not to harmonise the standards of the Parties. The Parties are determined to maintain and improve their respective high standards in the areas covered by this Title.

This approach is elaborated in Article 387.2 which states that parties shall not weaken or reduce their social or labour protections in a manner affecting investment between the parties from where the standards are at the end of the Brexit Transition Period (the end of 2020). This requirement applies also to environmental protection under Article 391.2

Chapter 3 covers the LPF obligation as it applies to subsidies. The agreement enables unilateral retaliation against illegal subsidies without the prior authorisation of an arbitration tribunal, much as under the World Trade Organization (WTO) Agreement on Subsidies and Countervailing Measures (SCM). The threshold for such retaliation through the removal of tariff concessions is high – it requires “serious risk that it will cause, a significant negative effect on trade or investment between the Parties.” Negative effects must be “based on facts and not merely on allegation, conjecture or remote possibility.” The “the change in circumstances … create a situation in which the subsidy would cause such a significant negative effect must be clearly predictable.”[5] This appears to contemplate a blacklist by which the TCA’s Partnership Council, the joint body charged with managing the implementation of the agreement, would identify subsidy types that would fall on a per se basis into this category. The language evidently suggests a rather high threshold for the retaliation to be warranted.

There is an additional specialized ‘rebalancing’ section in the TCA designed to respond to significant divergences between the parties in labour, the environment and subsidy policy. The essence is contained in paragraph 411.2:

If material impacts on trade or investment between the Parties are arising as a result of significant divergences between the Parties in [labour and social, environmental or climate protection, or with respect to subsidy control], either Party may take appropriate rebalancing measures to address the situation. Such measures shall be restricted with respect to their scope and duration to what is strictly necessary and proportionate in order to remedy the situation. Priority shall be given to such measures as will least disturb the functioning of this Agreement. A Party’s assessment of these impacts shall be based on reliable evidence and not merely on conjecture or remote possibility.

Such ‘non-regression’ clauses (also known as ‘standstill’ or ‘ratchet’ clauses) are common in modern FTAs. They are seen as a key part of the movement towards the recognition of the role of states in ensuring the observation by investors of corporate social responsibility principles, such as they relate to the environment as well as human rights.[6]

Such clauses are thought to have had provided rather limited protection, at least in the case of the environment.[7] There have been very few disputes arising under FTAs for breach of non-regression provisions. One such claim was brought under the Dominican Republic-Central America United States Free Trade Agreement (CAFTA-DR). The US argued that Guatemala had failed to enforce its own labour laws in part by preventing workers from forming unions, allegedly resulting in their competitive advantage over US workers. But the tribunal was unable to find that any weakening took place in a manner affecting trade, ruling that “attempting to establish that an effect on prices is due to a failure to enforce and not to … other factors would often be so fraught with difficulty as to make proof of trade effects impossible.”[8] This assessment was despite the fact that the CAFTA tribunal articulated a relatively weak trade effects test; the US had needed to demonstrate merely that Guatemala’s disputed practices had conferred “some competitive advantage on an employer or employers engaged in trade with the United States.”[9] Commentators have noted that this appears to be a fairly low threshold, compared for example to the injury test in safeguard proceedings, which requires substantial industry-wide effects.[10] Yet the CAFTA tribunal still found no evidence that any cost savings that might have accrued to Guatemalan exporters as a result of the alleged enforcement failures provided a competitive advantage. Non-regression clauses would seem to impose rather limited obligations on parties.

Under the TCA the threshold for harm which breaches the LPF obligation is undoubtedly a high one, as evident in the adjective ‘material.’ Questions of interpretation under the TCA are to be resolved by reference to public international law (not EU or UK law), as provided under Title II Article 4.1.

In public international law a ‘material breach’ of a treaty is one which violates ‘a provision essential to the accomplishment of the object or purpose of the treaty.’[11] It contemplates a violation so serious that it allows the other party to terminate the treaty. This will not be an easy standard to meet. The phrase ‘material impact’ of Article 411.2 differs from material breach but still calls for an interpretation based on the ordinary meaning of the words.[12] The English dictionary definition of ‘material’ is one which contemplates a high threshold, synonymous with that which is ‘significant, important’ or ‘significant or relevant, especially to the extent of determining a cause or affecting a judgement.’[13] This suggests that only the most egregious impacts on trade or investment are likely to be caught by this provision.

              The constrained nature of the LPF rebalancing system is also revealed by an understanding of the meaning of ‘impact’ on investment. The dictionary definition of ‘impact’ speaks of ‘a strong effect,’[14] suggesting something momentous as opposed to mundane, an outcome which is further emphasized by the modifier ‘material.’ The principal of effectiveness in treaty interpretation establishes that any provision in a treaty must be understood to have some significance and to achieve some end – it cannot be interpreted in such a way that would render the word meaningless.[15] Accordingly the word ‘material’ is not redundant – it an adjective which confines the notion of impact to the most serious cases. The scope for reaction against mis-alignment between the two parties appears even narrower in this light.

Marked changes in comparative advantage, suggesting long-term transformations in trade or investment flows, will not be easy to demonstrate let alone attribute to certain regulatory causes, as the tribunal in the US v Guatemala case observed. This is because Foreign Direct Investment (FDI) flows tend to fluctuate significantly from year to year in response to many different factors.[16] It would accordingly be difficult to discern the signal of impacts from lower standards from the noise of movements from all other causes. Trade flows are somewhat more stable, but they are also highly variable, as well as dependent on other factors (such as the Covid-19 epidemic).[17] Parsing out which changes have resulted from which legal interventions, rather than other systemic or random factors, would be a difficult task.

Even more problematic, an assessment of the impacts for breach of the LPF obligations must be based on ‘reliable evidence and not merely on conjecture or remote possibility.’ But, again, in most cases, establishing a counter-factual – the situation that would have occurred had there not been material regulatory divergence on the LPF – would be hard without extensive guesswork.[18] For example, in the case of a diversion in foreign investment, the complainant would need to show that an investor that would have located in its territory decided instead to establish in the other party’s territory because the regulatory environment in that jurisdiction was more attractive, not to mention lower or weaker in some measurable way. Alternatively, the complainant would need to show that the lower regulation in the other party’s territory was the reason that an investor moved from its former location in one party into the other party’s territory. In either case, establishing an investor’s strategic motivation in this way would be a tough evidentiary burden.[19] The fact that the TCA neglected to provide a definition for ‘significant divergences’ further indicates that this test will be hard to apply in practice.[20]

Most commentators appear to agree that the retaliation against departures from the LPF obligations will be difficult to challenge. One asserts ‘in practice it may not be all that easy to impose these [retaliation] measures, and if that’s the case, there won’t be much impact on a government’s regulatory decisions.’[21] Another notes “having to prove that a lowering of labour or environmental protection actually has trade or investment effects may appear so formidable a condition as to render the disciplines of the non-regression and non-enforcement clauses illusory.”[22] The assessment of what is meant by lowering of a standard in terms of an associated impact on the environment is a major problem with non-regression clauses in FTAs, severely impairing their functionality. It has been said that ‘[t]he difficulties and uncertainties in forecasting and measuring the effectiveness of environmental policies … as it develops over time create unanswered problems in applying non-regression clauses in practice.’[23] Environmental regulations pose particular challenges for trade tribunals[24] precisely because their causes and results are often diffuse and scientifically complex, complicating evidence-based approaches to establish sufficient ‘equivalence’ between approaches.

Some commentators believe that the TCA’s rebalancing tests are broad in some respects. This is because they cover any weakening or reduction in domestic levels of labour and environmental protection, not just waivers or derogations from laws, or sustained or recurring non-enforcement.[25] Another commentator similarly writes: ‘The process is startling because of … the lax test for when they may be imposed …[it] requires there to be a “material impact” before these can be imposed but it is not clear what the word “material” adds. Any impact will have material effects.’[26] But the notion that divergences are presumptively impactful is not credible. If the qualifier ‘material’ did not expand the understanding of impact, then it would not have been included in the text of the agreement at all, under the principle of effectiveness mentioned earlier. Others hold that the rebalancing mechanism ‘gives the parties significant room of manoeuvre to self-judge whether a certain situation justifies unilateral corrective measures.’ The LPF provisions in the TCA ‘could be an invitation to threaten to trigger these mechanisms, which in turn could provoke numerous small-scale trade wars.’[27] Yet again, his claim of ‘significant room to manoeuvre’ does not appear tenable given the clearly exceptional nature of the legal tests for materiality and necessity, as these concepts are understood in international law.

The primary remedy for breach of the TCA is to bring the offending measure into conformity with the agreement’s obligations. Tariffs are presented as a secondary remedy. The agreement therefore manages to preserve the parties’ right to pursue an autonomous approach to regulation, subject to potential retaliatory tariffs. Retaliation can apply across different areas of the agreement,[28] permitting the EU and the UK to choose the most politically sensitive areas against which to take retaliatory steps.

Yet again, the ‘strictly necessary’ standard for measures taken in response to material deviations in the LPF offers limited scope for the complainant to impose any such tariffs. Under public international law the standard of ‘necessity’ is usually taken to mean that it is ‘the only means for the State to safeguard an essential interest against a grave an immanent peril.’[29] In the field of trade, ‘necessity’ under the WTO’s General Agreement on Tariffs and Trade (GATT) Article XX General Exceptions is similarly confined. It is understood to mean something close to indispensable; that there is no less trade restrictive way to achieve the desired goal.[30] Likewise, investment jurisprudence suggests that this is a very tough test for host states to meet.[31]

The TCA’s LPF rebalancing tariffs must further be ‘proportionate.’ Proportionality tests are believed to be evident in the decisions of WTO panel and international investment law tribunals.[32] This exercise may be likened to the assessment of appropriate countervailing duties against subsidies or even the quantum of compensation payable in international investment arbitration for expropriations. In the case of the LPF rebalancing, calibrating appropriate tariffs in response to material impactful divergences would be complex and complicated.  The lack of a clear definition of appropriate “rebalancing measures” ideally with examples, further muddles the assessment.[33]

Article 761.3 of the TCA provides details on remedies for the purposes of the LPF provisions. One of the more significant rules is as follows: “The level of nullified or impaired benefits requested by the complaining Party or determined by the arbitration tribunal: (a) shall not include punitive damages, interest or hypothetical losses of profits or business opportunities.”[34] While punitive damages are conventionally excluded from remedies as a principle of international law[35], the reference to hypothetical losses underscores the need for clear evidence-based harms to trade or investment flows.

Even were it feasible for a party to quantify the impact on trade or investment flows, this would need to be expressed not in monetary damages (as in the case of remedies in international investment law) but translated into preferential tariff concessions in favour of the other party which would thereby be removed, more analogous to countervailing duties in trade law. While declines in trade flows could provide some guidance for retaliatory tariffs on goods, for investment impact, as noted above there would need to be a calibration of the harmful effect of the investor’s movement from one jurisdiction to the other, or even more problematically, its failure to establish in one party in the first place, instead choosing the other party’s territory, in which case the ‘impact’ would be entirely hypothetical. A plausible guide to compensation in the latter case could be situations in which pre-establishment national treatment was breached.[36] This might involve re-location costs or wasted capital expenses on premises and so on – a very messy exercise.

In addition to main remedy of tariffs for departures from the LPF, there is also the possibility for a limited reopening of the LPF commitments, providing essentially for a long-term rebalancing system aimed at accommodating permanent regulatory divergence. Overuse of the rebalancing system (whatever that might mean) can trigger a review of the LPF commitments entirely.[37] For example, if the UK pursued more far-reaching environmental policies it may argue that EU business had an unfair advantage. Under such circumstances, it may seek to activate the ‘rebalancing’ mechanism in the short term and apply tariffs accordingly. In the longer term, the UK may wish to amend the TCA to take this divergence of approach into account.

7